Patterns of Surgical Care and Complications in Elderly Adults

Stacie Deiner, MD, MS; Benjamin Westlake, BS; Richard P. Dutton, MD, MBA


J Am Geriatr Soc. 2014;62(5):829-835. 

In This Article

Abstract and Introduction


Objectives. To determine whether procedures, hospitals visited, and complications would differ according to decade in elderly adults and from those of younger adults.

Design. Retrospective cohort study.

Setting. The Anesthesia Quality Institute National Anesthesia Clinical Outcomes Registry (NACOR) is the largest database of anesthesia cases from academic and community hospitals and includes all insurance and facility types across the United States.

Participants. Eight million six hundred thirty-two thousand nine hundred seventy-nine cases from January 2010 to March 2013 were acquired. After exclusion of individuals younger than 18, nonapplicable locations, and brain death, 2,851,114 remained and were placed into age categories (18–64, 65–69, 70–79, 80–89, ≥90).

Measurements. Participant, surgical, anesthetic, and hospital descriptors and short-term outcomes (major complications, mortality at <48 hours).

Results. The largest number of older adults had surgery in medium-sized community hospitals. The oldest age group (≥90) underwent the smallest range of procedures; hip fracture, hip replacement, and cataract procedures accounted for more than 35% of all surgeries. Younger old adults underwent these procedures plus a significant proportion of spinal fusion, cholecystectomy, and knee surgery. Older adults had greater mortality and more complications than younger adults. Participants undergoing exploratory laparotomy had the greatest likelihood of death in any age category except 90 and older, in which small bowel resection predominated. The proportion of emergency surgery and the mortality associated with emergency surgery was 30% higher in the oldest group (≥90) than in adults aged 18 to 64.

Conclusion. This article reports the pattern of surgical procedures, complications, and mortality found in NACOR, which is one of the few data sets that contains data from community hospitals and individuals with all types of insurance. Because the outcomes portion of the data set is under development, it is not possible to investigate the relationship between hospital type and complications or mortality, but this study underscores the magnitude of geriatric surgery that occurs in community hospitals as an area for future outcomes studies.


Older adults undergoing surgery represent a large proportion of the overall surgical population; information gathered from the National Hospital Discharge Survey reported that, in 2006, individuals aged 65 and older accounted for 35.3% of all inpatient procedures and 32.2% of all outpatient procedures,[1,2] but there is a paucity of scientific literature examining perioperative healthcare patterns in the oldest-old adults (≥75) despite their high risk of postoperative complications and mortality. Single- and even multicenter studies often have inadequate sample sizes to describe this surgical population in depth.[3] The magnitude and risk of surgery for older adults underscores the importance of identifying high-volume surgical procedures and healthcare delivery systems on which to focus perioperative outcomes studies in the future.

Research has addressed individual-level outcomes in elderly adults such as cardiac risk stratification, delirium assessment, postoperative cognitive dysfunction, frailty, and pneumonia,[4–8] but from the policy and planning perspective, it is also important to understand whether these risk factors are ameliorated or worsened by where elderly adults have surgery and what procedures they undergo.[9] Although some of the risk is due to the physiology of aging (e.g., decreased cardiac reserve) or a composite of conditions (frailty), the perioperative risk of elderly adults is a more-complicated question including factors that may not be amenable to intervention (e.g., genetics) and those that may be due to regional variation or resource-based clinical decisions.[10–13]

An example of an important epidemiological question for older adults undergoing surgery was raised from a study that examined the incidence of surgical procedures in the year before death.[14] The authors found that end-of-life "surgical intensity" varied significantly according to region of country and age after adjustment for comorbidity. In addition, age was inversely proportional to number of procedures in the final year of life, which the authors commented might mean that providers' thresholds for providing intervention may change with age. This study opened the discussion of what is appropriate care for dying elderly adults and whether people who receive surgery may be a function of where they live.

The epidemiology of geriatric surgical and anesthetic care is difficult to define in the absence of a specialized surveillance system; in anesthesiology, this is beginning to change. Multicenter anesthesiology outcomes research consortiums including the Outcomes Research Consortium at the Cleveland Clinic and the Multicenter Perioperative Outcomes Group (MPOG) have collected data and produced important findings on a variety of perioperative issues.[15–19] In 2010, the Anesthesia Quality Institute (AQI) created the National Anesthesia Clinical Outcomes Registry (NACOR) with support from the American Society of Anesthesiologists. This database is unique because it is the largest data set of clinical anesthesia cases and because it includes data from academic and private hospitals in all of the U.S. Census regions.[20] NACOR contains more than 10 million anesthetic records across age groups, insurance type, and facility type harvested from electronic billing and clinical data to capture practice and patient profiles. The database has the potential to link facility-level information with individuals-level preoperative risk factors, intraoperative events, and postoperative complications. This study used NACOR to define the demographic characteristics and outcomes of older adults undergoing surgery.

This was the first study using the NACOR database to compare the distribution of cases and outcomes in five age categories (18–64, 65–69, 70–79, 80–89, ≥90). The hypothesis was that the procedures, patterns of care, and outcomes would be different between age groups. Furthermore, it was thought that systematic differences in how older adults are cared for would be of use to policy-makers, funding institutes, and researchers seeking areas where improvement in care could have the largest effect.